This is from I. M. Gelfand's Algebra book.

Fractions $\displaystyle\frac{a}{b}$ and $\displaystyle\frac{c}{d}$ are called neighbor fractions if their difference $\displaystyle\frac{ad - bc}{bd}$ has numerator of $\pm 1$, that is, $ad - bc = \pm 1$. Prove that

(a) in this case neither fraction can be simplified (that is, neither has any common factors in numerator and denominator);

(b) if $\displaystyle\frac{a}{b}$ and $\displaystyle\frac{c}{d}$ are neighbor fractions, then $\displaystyle\frac{a+c}{b+d}$ is between them and is a neighbor fraction for both $\displaystyle\frac{a}{b}$ and $\displaystyle\frac{c}{d}$; moreover,

(c) no fraction $\displaystyle\frac{e}{f}$ with positive integer $e$ and $f$ such that $f < b + d$ is between $\displaystyle\frac{a}{b}$ and $\displaystyle\frac{c}{d}$.

Parts (a) and (b) weren't too difficult, but I'm stuck on part (c). I've included (a) and (b) in case they're related to the solution to (c).

  • 7
    $\begingroup$ Look up "Farey sequence". $\endgroup$ – Yuval Filmus Jun 5 '11 at 19:05
  • $\begingroup$ Okay, I'll do that. Thanks! $\endgroup$ – Matt Gregory Jun 5 '11 at 19:09

Assume $\frac{e}{f}$ is (strictly) between $\frac{a}{b}$ and $\frac{c}{d}$. Then $\left|\frac{a}{b}-\frac{e}{f}\right| + \left|\frac{e}{f}-\frac{c}{d}\right| = \left|\frac{a}{b}-\frac{c}{d}\right| = \frac{1}{bd}$

But $\left|\frac{a}{b}-\frac{e}{f}\right| \geq \frac{1}{bf}$ and $\left|\frac{e}{f}-\frac{c}{d}\right|\geq \frac{1}{df}$. So $\frac{1}{bf} + \frac{1}{df} \leq \frac{1}{bd}$. Multiply both sides by $bdf$ and we get that $b+d\leq f$.

  • 1
    $\begingroup$ There is a geometric meaning to this, which is related to lattice points on the plane. If $(a,b)$ and $(c,d)$ are points on the plane, then $|ad-bc|$ is the area of the quadrilateral with points $(0,0)$, $(a,b)$, $(c,d)$ and $(a+c,b+d)$. If $a,b,c,d$ are integers, and $|ad-bc| = 1$, this theorem essentially says that there is no lattice point in the inside of the quadrilateral. $\endgroup$ – Thomas Andrews Jun 5 '11 at 20:26
  • $\begingroup$ To be clear, what does "between" mean in this context? Can $\frac{e}{f}$ be equal to one of the other fractions? $\endgroup$ – MathematicsStudent1122 Jul 8 '15 at 7:49
  • $\begingroup$ In this case, I wrote "strictly' precisely because I wanted to be clear that I meant $\frac{a}{b}<\frac{e}{f}<\frac{c}{d}$ or $\frac{c}{d}<\frac{e}{f}<\frac{a}{b}$. @ArshJhaj The theorem is not true if you allow $e=c,f=d$. The first part of my line is true for $\frac{e}{f}=\frac{a}{b}$, but the next step: $$\left|\frac{a}{b}-\frac{e}{f}\right|\geq \frac{1}{bf}$$ is obviously not $\endgroup$ – Thomas Andrews Jul 8 '15 at 10:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.